6. AOACSPIFANMethods-2018Awards

215

Zywicki & Sullivan: J ournal of AOAC I nternational V ol. 98, N o. 5, 2015  1415

0.250, 0.500, 1.00, 10.0, 50.0, and 100 ppb calibration standard curve points, a 5.00 ppb was added and the 100 ppb was deleted resulting in 0.250, 0.500, 1.00, 5.00, 10.0, and 50.0 ppb points. One participant mentioned issues with RSDs and IS drift when the method had not been performed on their instrument for a period of time but commented that adequate conditioning resolved the issues. The Study Director thoroughly reviewed all deviations and was confident, based on an overall assessment of the QC check information provided and statistical analysis of the results, that no impact to the data was evident. All of the laboratories’ results are presented in Table 2. Table  2012.15A shows the statistical evaluations for all the samples analyzed in this multilaboratory testing study. The RSD r ranged from 0.77 to 4.78%, and the RSD R ranged from 5.42 to 11.5%. The HorRat values for all results ranged from 0.35 to 1.31%. Repeatability and reproducibility for all seven samples were below the limits set forth in AOAC SMPR 2012.008 (4). All 13 laboratories’ data were included for statistical analysis for both RTF samples. Outliers for the powdered reconstituted samples and NIST SRM 1849a were removed prior to performing statistical analysis based on Cochran’s and Grubbs’ outlier tests. Upon completion of the collaborative study, comparison of data for the reconstituted powders revealed five laboratories’ results (Laboratories C, E, H, I, and L) were approximately 9 to 10 times higher than the other eight laboratories’ data. The other eight laboratories’ data agreed with values obtained during the SLV. The consistent factor of 9 to 10 suggested a calculation error, which agreed with the reconstitution factor (e.g., 225 g ÷ 25 g = 9). After correspondence with the five laboratories whose data were in question, it was evident that a misunderstanding of the calculation requirements for the reconstituted powders had occurred. The five laboratories had calculated the reconstituted powdered sample results on a dry basis instead of on an “as fed” basis. Laboratories C, I, and L submitted recalculated results prior to the collaborative study report submission due date, allowing inclusion of their data in the results table. Laboratory H submitted acceptable data but only after the due date. Laboratory E did not submit recalculated data. Since laboratories E and H recalculated reconstituted powder data were not received in time to include in the report, their original data were reported. Several comments to strengthen the method were provided during the SPIFAN ERP meeting in March 2015: Clarify in the method that it is not applicable to samples containing FD&C Red Dye No. 3 (erythrosine). Point out the possible need for increased instrument maintenance when using the method. Include precautions about the lens and/or lens stack possibly requiring additional maintenance and that analysis would benefit from thoroughly conditioning the instrument. Clarify the use and/or preparation of second source standards for CCV standard solutions. If acidic sample matrixes are typically analyzed on the ICP-MS instrument, perform a thorough cleaning of the entire sample introduction system and appropriate conditioning prior to analyzing basic matrixes. Clarify the importance of adhering to the peristaltic pump tubing sizes recommended for introducing IS and carrier solutions. If possible, maintain a dedicated set of cones and/or lens.

These suggestions have been accepted and incorporated into the method. Also incorporated were minor modifications taken from comments provided by several collaborators as well as incorporation of components requiring clarification as suggested by the Study Director. The overall results demonstrated that the method is fit- for-purpose to determine iodine levels in infant formula and adult/pediatric nutritional formula, and the Study Director recommended that it be adopted Official Final Action. It was the recommendation of the Study Director that the method is fit-for-purpose in determining total iodine in infant formula and adult/pediatric nutritional formula by ICP-MS and that it be adopted as an AOAC Final Action Official Method. The AOAC ERP evaluated the data presented in the final report for the collaborative study of AOAC First Action Official Method 2012.15 in March 2015 after which the method was recommended Final Action status. Subsequently, the Official Methods Board approved the method for Final Action in June 2015. We are very grateful to all laboratories that participated in the study, especially to the following collaborators (listed alphabetically based on their institution name): Angela Song and Fang Tian (Abbott Nutrition International ANRD, China) Kathryn Stanley and Mark W. Collison (Archer Daniels Midland Co., Decatur, IL) Grant Fulford and Pee Yang (Covance, Madison, WI) Yirong Xu, Sook Yain Chen, and KienHeng Lee (Covance, Singapore) Marvin Boyd Jr, and Cheryl D. Stephenson (Eurofins Central Analytical Laboratories, New Orleans, LA) Josh Messerly (Eurofins Nutrition Analysis Center, Des Moines, IA) Anders K. Svaneborg and Per K.B. Nilsson (Eurofins Steins LaboratoriumA-S, Denmark) Sudhakar Yadlapalli (First Source Laboratory Solutions LLP, Hyderabad, India) Cai Weihong [Guangzhou Quality Supervision and Testing Institute (GZ GQT), China] Raymond Yu (National Center of Supervision and Inspection on Food Stuff Product Quality, Shanghai, China) Xiaojun Deng (Shanghai Exit and Entry Inspection and Quarantine Bureau, China) Wu Bolong (Test Center of ChineseAcademy of Inspection and Quarantine, China) Qin Xu and Lei Bao (The Food and Agricultural Products Testing Agency of Shandong CIQ, China). A debt of gratitude goes out to Meagan (Dallman) Knautz, Covance Laboratories, Inc., Madison, WI, for her significant contributions during method development and leading up to and including the SLV. A sincere thank you goes to Darryl Sullivan, Covance Laboratories, Inc., Madison, WI; Robert Rankin, Infant Nutrition Council of America, Washington, DC; and James Wehrmann, Covance Laboratories, Inc., Madison, WI, for their guidance throughout the entire process. Special thanks Recommendations Acknowledgments